sign in sign up
<<
Home , 2018 BD, GJ 9827, HD 219134 b, HD 3167, LHS 1140, LHS 3844 b

A&A 636, A58 (2020) https://doi.org/10.1051/0004-6361/201937179 & © ESO 2020 Astrophysics

A hot terrestrial orbiting the bright M dwarf L 168-9 unveiled by TESS ?,??,??? N. Astudillo-Defru1, R. Cloutier2,3, S. X. Wang4, J. Teske4,????, R. Brahm5,6,7, C. Hellier8, G. Ricker9, R. Vanderspek9, D. Latham2, S. Seager10, J. N. Winn11, J. M. Jenkins12, K. A. Collins2, K. G. Stassun13, C. Ziegler14, J. M. Almenara15, D. R. Anderson8,16, E. Artigau17, X. Bonfils15, F. Bouchy18, C. Briceño19, R. P. Butler20, D. Charbonneau2, D. M. Conti21, J. Crane4, I. J. M. Crossfield22,23, M. Davies12, X. Delfosse15, R. F. Díaz24,25, R. Doyon17, D. Dragomir26, J. D. Eastman2, N. Espinoza27, Z. Essack10, F. Feng20, P. Figueira28,41, T. Forveille15, T. Gan30, A. Glidden9, N. Guerrero9, R. Hart31, Th. Henning32, E. P. Horch33, G. Isopi34, J. S. Jenkins35, A. Jordán36,7, J. F. Kielkopf37, N. Law38, C. Lovis18, F. Mallia34, A. W. Mann38, J. R. de Medeiros39, C. Melo28, R. E. Mennickent29, L. Mignon15, F. Murgas15, D. A. Nusdeo40, F. Pepe18, H. M. Relles2, M. Rose12, N. C. Santos41,42, D. Ségransan18, S. Shectman4, A. Shporer9, J. C. Smith43,12, P. Torres5, S. Udry18, J. Villasenor44, J .G. Winters2, and G. Zhou2 (Affiliations can be found after the references)

Received 22 November 2019 / Accepted 27 January 2020

ABSTRACT

We report the detection of a transiting super--sized planet (R = 1.39 0.09 R ) in a 1.4- orbit around L 168-9 (TOI-134), a bright M1V dwarf (V = 11, K = 7.1) located at 25.15 0.02 pc. The± host ⊕ was observed in the first sector of the Tran- siting Survey Satellite (TESS) mission. For confirmation± and planet measurement purposes, this was followed up with ground-based , seeing-limited and high-resolution imaging, and precise (PRV) observations using the HARPS and Magellan/PFS spectrographs. By combining the TESS data and PRV observations, we find the mass of L 168-9 b to be +0.44 4.60 0.56 M and thus the bulk to be 1.74 0.33 times higher than that of the Earth. The is smaller than 0.21± (95% confidence).⊕ This planet is a level one candidate− for the TESS mission’s scientific objective of measuring the of 50 small , and it is one of the most observationally accessible terrestrial planets for future atmospheric characterization. Key words. : individual: L 168-9 – planetary systems – stars: late-type – techniques: photometric – techniques: radial velocities

1. Introduction Telescope (ELT, de Zeeuw et al. 2014) will have unprecedented capabilities in performing detailed studies of the atmospheres Transiting planets are the best planet candidates for perform- of terrestrial planets, and the interpretation of the results will ing detailed characterizations, first and foremost, because they require an accurate mass measurement (e.g., Batalha et al. allow for the possibility of unambiguous mass measurements 2019). (e.g., HD 209458b, Charbonneau et al. 2000; Mazeh et al. 2000; Scientific operations in the Transiting Exoplanet Survey Henry et al. 2000). We can determine the minimum mass of Satellite (TESS; Ricker et al. 2015) started in July 2018; the aim the planet (m sin i) from the Doppler effect and also the plane- of which is to detect transiting planets around bright and nearby tary radius and the (i) from the light stars that are bright enough for Doppler mass measurements. For curve, thus yielding a measurement of the planet’s mass. More- this task, TESS surveys about 85% of the sky during the prime over, from this combination, we can calculate the planet’s mean mission. The survey covering the southern hemisphere is density and shed light on its internal structure by comparing it now complete, and the northern survey is underway. Each hemi- with models containing different amount of iron, silicates, water, sphere is divided into 13 rectangular sectors of 96◦ 24◦ each. hydrogen, and helium. Furthermore, transiting planets are unique Each sector is continuously observed for an interval× of 27-days, because of the feasibility in characterizing the upper atmosphere with a cadence of 2 min for several hundreds of thousands of by spectroscopy during transits and occultations of a signifi- preselected stars deemed best suited for planet searching. Addi- cant number of planets. The forthcoming James Webb Space tionally, during the TESS prime mission, the full frame images, Telescope (JWST, Gardner et al. 2006) and the Extremely Large which consist of the full set of all science and collateral pixels across all CCDs of a given camera, are available with a cadence ? Full Tables B.1 and B.2 are only available at the CDS via anony- of 30 min. M dwarfs are of special interest because the transit mous ftp to cdsarc.-strasbg.fr (130.79.128.5) or via http:// and radial-velocity signals of a given type of planet are larger cdsarc.u-strasbg.fr/viz-bin/cat/J/A+A/636/A58. for these low-mass stars than they are for -like stars. In addi- ?? Partially based on observations made with the HARPS instrument on the ESO 3.6 m telescope under the program IDs 198.C-0838(A), tion, compared to hotter stars, M dwarfs present better conditions 0101.C-0510(C), and 1102.C-0339(A) at Cerro La Silla (Chile). for the detection of planets orbiting the circumstellar habitable ??? This paper includes data gathered with the 6.5 m Magellan zone: performing surveys in M dwarfs is less time consuming, Telescopes located at Las Campanas Observatory, Chile. and there are larger Doppler signals and an increased transit ???? NASA Hubble Fellow. probability. Sullivan et al.(2015) anticipates that out of the 556

Article published by EDP Sciences A58, page 1 of 13 A&A 636, A58 (2020) small (<2 R ) transiting planets discovered by TESS, 23% of Table 1. L 168-9 (TIC 234994474) properties. ⊕ them will be detected orbiting bright (KS < 9) stars and that 75% of small planets will be found around M dwarfs. Parameter Units Value Reference This paper reports the discovery of a small planet orbiting the star L 168-9 (TOI-134), based on TESS data. The host star RA [J2000] 23h20m07.52s Gaia2018 is a bright M dwarf. An intense precise radial-velocity cam- Dec [J2000] 60◦03054.6400 Gaia2018 paign with HARPS and the Planet Finder Spectrograph (PFS) Spectral type− M1V Ga2014 has revealed the terrestrial nature of the newly detected world. B [mag] 12.45 0.19 Ho2000 This work is presented as follows: Sect.2 describes the host V [mag] 11.02 ±0.06 Ho2000 star properties. Sections 3.1 and 3.3 describe the photometric B [mag] 12.460± 0.025 He2016 A ± and radial-velocity observations. Section4 presents an analy- VA [mag] 11.005 0.018 He2016 sis of all the data, including the study of stellar activity. Finally, g [mag] 11.752 ±0.032 He2016 A ± Sect.6 places L 168-9 b within the larger context of the sample RA [mag] 10.416 0.028 He2016 of detected planets. ± iA [mag] 9.675 He2016 W [mag] 6.928 0.060 Cu2013 1 ± 2. L 168-9 W2 [mag] 6.984 0.020 Cu2013 W [mag] 6.906± 0.016 Cu2013 L 168-9, also known as CD-60 8051, HIP 115211, 2MASS 3 W [mag] 6.897 ± 0.074 Cu2013 J23200751-6003545, with the entry 234994474 of the TESS 4 T 9.2298 ± 0.0073 St2018 input catalog (TIC) or 134 of the TESS object of interest (TOI) ± list, is a of spectral type M1V. It appears in the J 7.941 0.019 Cu2003 H 7.320 ± 0.053 Cu2003 southern sky and resides at a distance of 25.150 0.024 pc ± from the Sun (Gaidos et al. 2014; Collaboration± 2018). Ks 7.082 0.031 Cu2003 B 11.2811± 0.0016 Gaia2018 Table1 lists the key parameters of the star, namely its posi- p ± tion, visual and near-infrared apparent magnitudes, parallax, G 10.2316 0.0008 Gaia2018 R 9.2523 ±0.0011 Gaia2018 , secular acceleration, and its essential physical p ± properties. π [mas] 39.762 0.038 Gaia2018 Distance [pc] 25.150 ± 0.024 Gaia2018 We performed an analysis of the broadband spectral energy 1 ± distribution (SED) together with the Gaia DR2 parallax in order µα [mas yr− ] 319.96 0.10 Gaia2018 1 − ± to determine an empirical measurement of the stellar radius, fol- µδ [mas yr− ] 127.78 0.12 Gaia2018 dv /dt [m s 1 yr 1] 0.06865− ±0.00011 this work lowing the procedures described by Stassun & Torres(2016) r − − ± and Stassun et al.(2017, 2018a). We took the B V magnitudes Ms [M ] 0.62 0.03 Ma2019 T T ± from Tycho-2, the BVgri magnitudes from APASS, the JHKS Rs [R ] 0.600 0.022 Sect.2 T [K] 3800± 70 Sect.2 magnitudes from 2MASS, the W1–W4 magnitudes from WISE, eff ± and the G magnitude from Gaia. Together, the available pho- Ls [L ] 0.0673 0.0024 Sect.2 3 ± tometry spans the full stellar SED over the wavelength range of log(g) [g cm− ] 4.04 0.49 Sect.2 ± 0.35–22 µm (see Fig.1). [Fe/H] 0.04 0.17 Ne2014 ± We performed a fit using NextGen mod- log(R0 ) 4.562 0.043 As2017A HK − ± els (Hauschildt et al. 1999); the effective (T ) and Prot [days] 29.8 1.3 Sect. 4.1 eff ± (log g) were constrained on the ranges reported in the TIC (Stassun et al. 2018b), while the [Fe/H] References. Gaia2018 – Gaia Collaboration(2018); Ga2014 – Gaidos was fixed to a typical M-dwarf metallicity of 0.5. We fixed et al.(2014); Ho2000 – Høg et al.(2000); H22016 – Henden et al.(2016); − Cu2013 – Cutri et al.(2013); St2018 – Stassun et al.(2018b); Cu2003 – the extinction (AV) to zero, considering the proximity of the star and that the degrees-of-freedom of the fit is 10. The resulting fit Cutri et al.(2003); Ma2019 – Mann et al.(2019); Ne2014 – Neves et al. (2014); As2017a – Astudillo-Defru et al.(2017a). (Fig.1) has a χ2 of 42.3 (χ2 = 4.2), with T = 3800 70 K. red eff ± The relatively high χ2 is likely due to systematics since the stel- Moór et al. 2006) in which χ 3 represents a robust detection lar atmosphere model is not perfect. We artificially increased the λ of IR excess. We obtained χ =≥0.70 in the W band, ruling out observational uncertainty estimates until χ2 = 1 was achieved. λ 3 red the presence of a around L 168-9. Integrating the (unreddened) model SED gives the bolometric flux at Earth of F = 3.41 0.12 10 9 erg s 1 cm 2. By bol ± × − − − taking the Fbol and Teff together with the Gaia DR2 parallax, 3. Observations which was adjusted by +0.08 mas to account for the system- atic offset reported by Stassun & Torres(2018), this gives the The first hint of a planetary companion orbiting L 168-9 came stellar radius of R = 0.600 0.022 R . Finally, estimating the about following the analysis of TESS data. After the data valida- from the empirical± relations of Mann et al.(2019) tion report was released to the community, a follow-up campaign gives M = 0.62 0.03M . With these values for the mass and started with several instruments and was performed by differ- ± 3 radius, the stellar mean density is ρ = 4.04 0.49 g cm− . We ent teams to check on whether the transit-like signal seen by also tested to fit the SED by using a BT-Settl± theoretical grid TESS originated from a planet, as opposed to a stellar binary or of stellar model (Allard 2014), where we obtained a consistent other source. The follow-up observations included supplemen- result. tary time series photometry aiming to detect additional transits, We searched for infrared (IR) excess in WISE data using the seeing-limited and high-resolution imaging to analyze the possi- Virtual Observatory SED Analyser (VOSA, Bayo et al. 2008), bility that the signal comes from a star on a nearby sightline, and which could indicate the presence of debris disks. We computed precise radial-velocity monitoring to measure the companion’s the excess significance parameter (χλ, Beichman et al. 2006; mass.

A58, page 2 of 13 N. Astudillo-Defru et al.: L 168-9 b: a nearby hot terrestrial world

The TESS time series covers 19 transits for what was orig- -9 inally deemed a planet candidate (L 168-9 b or TOI-134.01), and it was reported on the TESS exoplanet follow-up observ- ) -2 ing program (TFOP) website5. According to the data validation -10

cm report, the is P = 1.401461 0.000137 days -1 and the transit depth is ∆F/F = 566 38 ppm,± which trans- 0 ± -11 lates into a planetary radius of Rp = 1.58 0.36 R (Sect.2 ± ⊕ (erg s

λ describes how we determined the stellar radius, which is the F

λ same value as the one used in the validation report). The time -12 of midtransit at an arbitrarily chosen reference is (BJD) log Tc = 2458326.0332 0.0015. -13 ± 3.1.2. LCOGT, MKO, and SSO T17 1 10 We acquired ground-based time series photometric follow-up of λ (µm) L 168-9 and the nearby field stars as part of the TESS follow- Fig. 1. Spectral energy distribution (SED) of L 168-9. Red error bars up observing program (TFOP) to attempt to rule out nearby represent the observed photometric measurements, where the horizontal eclipsing binaries (NEBs) in all stars that are bright enough to bars represent the effective width of the passband. Blue circles are the cause the TESS detection and that could be blended in the TESS model fluxes from the best-fit NextGen atmosphere model (black). aperture. We used the TESS Transit Finder6, which is a cus- tomized version of the Tapir software package (Jensen 2013), 3.1. Photometry to schedule our transit observations. We observed one full transit simultaneously by using three 1- 3.1.1. TESS m telescopes at the Las Cumbres Observatory Global Telescope (LCOGT; Brown et al. 2013) South Africa Astronomical Obser- TESS observed Sector 1 from 25 July to 22 August 20181, a vatory node on 21 September 2018 in the i0-band. The Sinistro 27.4-day interval that is typical of each sector. L 168-9 is listed detectors consist of 4K 4K 15-µm pixels with an image scale in the cool dwarf catalog, for which the known properties of 1 × of 000. 389 pixel− , resulting in a field-of-view of 26.05 26.05. as many dwarf stars as possible with V – J > 2.7 and effective The images were calibrated by the standard LCOGT BANZAI× lower than 4000 K were gathered (Muirhead et al. pipeline. 2018). The predicted TESS-band apparent magnitudes are also We observed a full transit from the Mount Kent Obser- given in this catalog. L 168-9 was chosen for 2-min time sam- vatory (MKO) 0.7-m telescope near Toowoomba, Australia on pling as part of the TESS candidate target list (CTL, Stassun 23 September 2018 in the r0-band. The Apogee U16 detector con- et al. 2018b), consisting of a subset of the TIC identified as sists of 4K 4K 9-µm pixels with an image scale of 0. 41 pixel 1, high-priority stars in the search for small transiting planets. × 00 − resulting in a field-of-view of 270 270. The images were cali- Time series observations of L 168-9 were made with CCD 2 of brated using the AstroImageJ (AIJ×) software package (Collins Camera 2. et al. 2017). The basic calibration, reduction, and detrending of the time We observed a full transit from the Siding Spring Observa- series as well as the search for transit-like signals were performed tory, Australia, iTelescope T17 0.43-m telescope on 27 Septem- in the TESS Science Processing Operations Center (SPOC) at ber 2018 with no filter. The FLI ProLine PL4710 detector consists the NASA Ames Research Center (Jenkins et al. 2016). The light of 1K 1K pixels with an image scale of 0. 92 pixel 1, resulting curves were derived by the SPOC pipeline, and they consist of × 00 − in a field-of-view of 15.05 15.05. The images were calibrated a time series based on simple aperture photometry (SAP) as using AstroImageJ. × well as a corrected time series based on presearch data condi- We used the AstroImageJ to extract differential light curves tioning (PDC, Smith et al. 2012; Stumpe et al. 2014) referred of L 168-9 and all known stars within 2.5 of the target star to as PDCSAP (as detailed in the TESS Science Data Products 0 2 that are bright enough to have possibly produced the shal- Description Document ). This work made use of the PDC- low TESS detection, which includes 11 neighbors brighter than SAP time series available on the Mikulski Archive for Space 3 TESS-band = 17.9 mag. This allows an extra 0.5 in delta mag- Telescopes (MAST ). nitude relative to L 168-9 to account for any inaccuracies in the A data validation report for L 168-9 was released to the com- 4 TESS band reported magnitudes in the TICv8. The L 168-9 light munity as part of the MIT TESS alerts . The report includes curves in all five photometric data sets show no significant detec- several of the following validation tests to assess the probability tion of the shallow TESS detected event, which is as expected that the signal is a false positive: eclipsing-binary discrimina- from our lower precision ground-based photometry. Considering tion tests, a statistical bootstrap test, a ghost diagnostic test, a combination of all five photometric data sets, we excluded all and difference-image centroid offset tests. These are described 11 known neighbors that are close enough and bright enough to by Twicken et al.(2018). All of the tests were performed L 168-9 to have possibly caused the TESS detection as potential successfully. The formal false-alarm probability of the planet sources of the TESS detection. candidate was reported to be 5.85 10 37. × − 1 3.1.3. WASP The Sector 1 pointing direction was RA(J2000):+352.68◦, Dec(J2000): 64.85◦, Roll:-137.85◦. WASP-South, located in Sutherland, South Africa, is the south- 2 − https://archive.stsci.edu/missions/tess/doc/ ern station of the Wide Angle Search for Planets (WASP, EXP-TESS-ARC-ICD-TM-0014.pdf 3 https://mast.stsci.edu/portal/Mashup/Clients/Mast/ 5 https://exofop.ipac.caltech.edu/tess/ Portal.html 6 https://astro.swarthmore.edu/telescope/tess-secure/ 4 https://tev.mit.edu find_tess_transits.cgi A58, page 3 of 13 A&A 636, A58 (2020)

0 TIC234994474 1 SOAR Speckle ACF 1 2 0 [arcsec] 3 -1 4 -1 0 1 [arcsec]

magnitude ( I -band) 5

6

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 arcsec

Fig. 2. SOAR speckle results of L 168-9. Black points represent the I-band contrast obtained at a given separation of the star. The solid black line shows the 5σ detection limit curve.

Pollacco et al. 2006). It consists of an array of eight cameras, each of which is backed by a 2048 2048 CCD. Observations in 2010 and 2011 (season A) used× 200 mm, f/1.8 lenses with a broadband filter spanning 400 700 nm and a plate scale of 1 − 13.700 pixel− . Then, from 2012 to 2014 (season B), WASP-South used 85 mm, f/1.2 lenses with a Sloan r’ filter and a plate scale of 1 3200 pixel− . The array rastered a number of fields during a clear night at typically 10-min cadence. Fig. 3. DSSI/Gemini-S detection limit curves of L 168-9 in the 692 nm L 168-9 was monitored for four consecutive , from (top) and 880 nm (bottom) filters. Squares and points on the left 20 May 2010 to 12 December 2014; typically covering 150 days represent local maxima and local minima, respectively. in each . In one campaign, two cameras with overlapping fields observed the star, giving a total of 27 300 data points; in another campaign, L 168-9 was observed by three cameras with South, Chile, on UT 28 October 2018 under program GS-2018B- overlapping fields, totaling 170 000 data points. The photometry LP-101 (PI: I. Crossfield). The usual 692 nm and 880 nm filters on DSSI were used, and three sequences of 60 ms/frame 1000 has a dispersion of 0.027 δmag and an average uncertainty × of 0.024 δmag, presenting clear signs of variability, as shown frames were taken. The total time on target, including read- below in Sect. 4.1. out overhead, was six minutes. Howell et al.(2011) and Horch et al.(2011, 2012) detail the speckle observing and data reduc- tion procedures. The detection limit curves are shown in Fig.3. 3.2. High-resolution imaging While the 692 nm image was taken at too low of a gain, leading to a shallow detection limit, the 880 nm detection limit curve The relatively large 2100 pixels of TESS can lead to photometric contamination from nearby sources. These must be accounted (b) in Fig.3 indicates that L 168-9 lacks any companions of ∆m 5.0 mag beyond 0.1 and any companions of ∆m 5.5 mag for to rule out astrophysical false positives, such as background ∼ 00 ∼ eclipsing binaries, and to correct the estimated planetary radius, beyond 0.200. Gemini-South and the SOAR result (Sect. 3.2.1) which was initially derived from the diluted transit in a blended means that the transit signal is likely to be associated with (Ziegler et al. 2018). Without this correction, the L 168-9. interpreted planet radius can be underestimated (e.g., Ciardi et al. Thus, we conclude that there is no significant contamination 2015; Teske et al. 2018). of the TESS photometric aperture that would bias the determina- tion of the planet radius. The Sinistro data (Sect. 3.1.2) rule out surrounding stars as potential sources of the TESS detection, so 3.2.1. SOAR we can assume the planet orbits L 168-9. Given the limits placed We searched for close companions to L 168-9 with speckle imag- on nearby companions by the Gemini-South data, if there were a close companion, it would have to be fainter by 5 magnitudes ing on the 4.1-m Southern Astrophysical Research telescope ∼ (SOAR, Tokovinin 2018) installed in Cerro Pachón, Chile, on than the primary star, meaning the planet radius correction factor would be at most 2018 September 25 UT using the I-band (λcen = 824 nm, full width at half maximum = 170 nm) centered approximately on p R = R 1 + 10 0.4∆m = R 1.005. (1) the TESS passband. Further details of the TESS SOAR survey p,corr p × − p × are published in Ziegler et al.(2020). This is smaller than the derived planet radius uncertainty Figure2 shows the 5 σ detection sensitivity. No nearby stars (Table2). (ρ < 100. 6) to L 168-9 were detected within the sensitivity limits of SOAR. 3.3. Radial velocity 3.2.2. Gemini-South 3.3.1. HARPS Observations of L 168-9 were conducted with the Differential The High Accuracy Radial velocity Planet Searcher (HARPS Speckle Survey Instrument (DSSI; Horch et al. 2009) on Gemini Mayor et al. 2003) is an echelle spectrograph that is mounted

A58, page 4 of 13 N. Astudillo-Defru et al.: L 168-9 b: a nearby hot terrestrial world

Table 2. Measured transit and RV model parameters of the L 168-9 .

Measured transit model parameters TESS Baseline flux, γ 0.00001 0.00029 TESS ± Orbital period, P [days] 1.40150 0.00018 b ± +0.00088 Time of midtransit, T0 [BJD-2,457,000] 1340.04781 0.00122 − Scaled semimajor axis, a/R 7.61 0.31 s ± Planet-star radius ratio, r /R 0.0212 0.001 p s ± +0.8 Orbital inclination, i [deg] 85.5 0.7 − +0.125 Linear limb darkening coefficient, q1 0.397 0.111 − +0.058 Quadratic limb darkening coefficient, q2 0.189 0.056 − +0.00011 TESS additive jitter, sTESS 0.00003 0.00003 − Radial velocity GP hyperparameters HARPS+PFS HARPS PFS 1 +1.38 +1.24 ln HARPS covariance amplitude, ln (aHARPS/m s− ) 3.27 1.03 3.24 1.19 - − − 1 +1.24 +1.23 ln PFS covariance amplitude, ln (aPFS/m s− ) 3.63 1.08 - 4.01 0.97 − − +2.81 +3.16 +2.02 ln RV exponential timescale, ln (λRV/day) 11.90 1.99 12.25 2.18 13.3 1.44 − − − +0.20 +0.41 +0.39 ln RV coherence, ln (ΓRV) 0.09 0.24 0.42 0.55 0.22 0.44 − − − − − − +0.02 +0.03 +0.04 ln RV periodic timescale, ln (PRV/day) 3.47 0.03 3.47 0.03 3.48 0.03 − − − 1 +0.20 +0.98 HARPS additive jitter, sHARPS [m s− ] 0.10 0.09 0.86 0.86 - − − PFS additive jitter, s [m s 1] 2.82 0.42 - 2.78 0.30 PFS − ± ± Measured RV model parameters HARPS+PFS HARPS PFS HARPS zero point velocity, γ [km s 1] 29.7687 0.0013 29.7692 0.0013 - HARPS − ± ± PFS zero point velocity, γ [km s 1] 0.00077 0.00186 - 0.00063 0.00142 PFS − ± ± 1 +0.47 +0.54 +0.52 Semiamplitude, K [m s− ] 3.66 0.46 3.74 0.59 3.26 0.70 − − − +0.17 +0.15 +0.18 h = √e cos ω 0.10 0.12 0.04 0.18 0.01 0.21 − − − − − +0.17 +0.22 +0.23 k = √e sin ω 0.00 0.16 0.09 0.23 0.01 0.30 − − − − Derived L 168-9 b parameters HARPS+PFS+TESS HARPS+TESS PFS+TESS Semimajor axis, a [AU] 0.02091 0.00024 ± Equilibrium temperature, Teq [K] Zero bond 965 20 ± -like bond albedo = 0.77 668 14 ± Planetary radius, Rp [R ] 1.39 0.09 ⊕ ± +0.71 +0.70 Planetary mass, Mp [M ] 4.60 0.56 4.74 0.75 4.08 0.90 ⊕ ± − − 3 +2.4 +2.5 +2.3 Planetary bulk density, ρp [g cm− ] 9.6 1.8 10.0 1.9 8.7 2.1 − − − 2 +4.5 +4.8 +4.4 Planetary surface gravity, gp [m s− ] 23.9 3.9 24.4 4.2 21.4 4.8 − − − 1 +1.4 +1.6 +1.6 Planetary escape velocity, vesc [km s− ] 20.5 1.4 20.8 1.7 19.4 2.2 − − − ( ) Orbital eccentricity, e † <0.21 <0.25 <0.26

( ) Notes. † 95% confidence interval. on the 3.6 m telescope at La Silla Observatory, Chile. The light or it can be placed on sky for moderate precision. HARPS is sta- is spread over two CCDs (pixel size 15 µm) by a science fiber bilized in pressure and temperature, and it has a resolving power and a calibration fiber. The calibration fiber can be illuminated of 115 000. The achievable precision in radial velocity is better 1 with the calibration lamp for the best radial velocity precision, than 1 m s− .

A58, page 5 of 13 A&A 636, A58 (2020)

technique (e.g., Baranne et al. 1996). Nevertheless, we derived radial velocities with a different approach to exploit the Doppler information of spectra as much as possible (e.g., Anglada- Escudé & Butler 2012). We performed a maximum likelihood analysis between a stellar template and each individual spec- trum following the procedure presented in Astudillo-Defru et al. (2017b). The stellar template corresponds to a true stellar spec- trum of the star in which the signal-to-noise ratio is enhanced. It was made from the median of all the spectra after shifting them into a common barycentric frame. The resulting template is Doppler shifted by a range of trial radial velocities to con- struct the maximum likelihood function from which we derived the HARPS radial velocity used in the subsequent analysis. The obtained radial velocities, which are listed in Table B.1, present 1 a dispersion of 4.01 m s− and a median photon uncertainty of 1 1.71 m s− .

3.3.2. PFS The Planet Finder Spectrograph is an iodine-calibrated, environ- mentally controlled high resolution PRV spectrograph (Crane et al. 2006, 2008, 2010). Since first light in October 2009, PFS has been running a long-term survey program to search for plan- ets around nearby stars (e.g., Teske et al. 2016). In January 2018, PFS was upgraded with a new large format CCD with 9µm pix- els and switched to a narrower slit for its regular operation mode to boost the resolution from 80 000 to 130 000. The PFS spectra were reduced and analyzed with a custom IDL pipeline that is 1 capable of delivering RVs with <1 m s− precision (Butler et al. 1996). We followed up L 168-9 with PFS on the 6.5 m Magellan II Clay telescope at Las Campanas Observatory in Chile from 13–26 October 2018 and then on 16 and 21 December 2018. Fig. 4. Left column: generalized Lomb-Scargle periodograms of the Observations were conducted on 15 nights, with multiple expo- HARPS and PFS RV time series, window functions, and the S-index, sures per night. There were a total of 76 exposures of 20 min Hα, Hβ, Hγ, and sodium doublet activity indicators. The vertical dotted each. We typically took 2–6 exposures per night over a range of lines highlight the locations of the L 168-9 b orbital period, the stellar timescales to increase the total S/N per epoch and also to average , and its first two harmonics. Right column: false alarm probabilities were computed from bootstrapping with replacement. out the stellar and instrumental jitter. Each exposure had a typical S/N of 28 per pixel near the peak of the blaze function, or 56 per 1 resolution element. The radial velocity dispersion is 4.61 m s− 1 We began monitoring L 168-9 with HARPS on 29 Septem- and the median RV uncertainty per exposure is about 1.8 m s− . ber 2018, which was soon after the TESS alert. We elected not to Five consecutive 20-min iodine-free exposures were obtained to use the simultaneous wavelength calibration (i.e., the on-sky cal- allow for the construction of a stellar spectral template in order ibration fiber) to ensure that the bluer spectral regions would not to extract the RVs. These were bracketed with spectra of rapidly be contaminated by the calibration lamp that provides a much rotating B stars taken through the iodine cell for reconstruction stronger flux for any instrumental setup. The exposure time was of the spectral line spread function and wavelength calibration set to 900 s for ESO programs 198.C-0838 and 1102.C-0339, for the template observations. and to 1200 s for ESO program 0101.C-0510, translating in a The PFS observations of L 168-9 that are presented here are median signal-to-noise ratio (S/N) per spectral pixel of 51 and part of the Magellan TESS survey (MTS) that will follow up on 70 at 650 nm, respectively. A single spectrum on 1 October 2018 30 super- and sub- discovered by TESS in the had an exposure time of 609 s for an unknown reason. A total of next∼ three years using PFS (Teske et al., in prep.). The goal of 47 HARPS spectra were collected, ending with observations on MTS is to conduct a statistically robust survey to understand the 19 December 2018. formation and evolution of super-Earths and sub-Neptunes. The HARPS spectra were acquired in roughly three packs of data observation schedules of all MTS targets, including L 168-9, can separated in time by about 40 days. This sampling is reflected be found on the ExoFOP-TESS website7. in the window function presented in Fig.4. Two archival spectra of L 168-9 are available at the ESO database. However, a radial velocity offset was introduced in May 2015 because the vacuum 4. Analysis vessel was opened during a fiber upgrade (Lo Curto et al. 2015). 4.1. Photometric rotation period As this offset is not yet well characterized for M dwarfs, we decided to disregard those two points (from July 2008 and June Knowledge of the period helps to disentangle spu- 2009) in our subsequent analysis. rious RV signals arising from rotation and true RV signals due The HARPS Data Reduction Software (Lovis & Pepe 2007) computes radial velocities by a cross-correlation function 7 https://exofop.ipac.caltech.edu/tess/

A58, page 6 of 13 N. Astudillo-Defru et al.: L 168-9 b: a nearby hot terrestrial world

Fig. 5. WASP photometry. Left column: photometry time series; the blue shaded regions depict 1σ about the mean GP regression model± of the binned photo- metry. Right column: generalized Lomb- Scargle periodogram of each season high- lights the prevalence of photometric varia- tions close to the measured rotation period and/or its first and second harmonics (ver- tical dashed lines). to orbital motion (e.g., Queloz et al. 2001; Cloutier et al. 2017). except for within the second WASP season wherein the dominant L 168-9 was photometrically monitored by WASP between May periodicity appears at the first harmonic of Prot 15 days. 2010 and December 2014 within five observing seasons, each Given periodicities that are significantly∼ detected in the lasting approximately 200 days in duration. The photometric pre- WASP photometry, we proceeded to measure the photometric cision, observing cadence, and baselines within the WASP fields rotation period of L 168-9 Prot with each WASP camera and in are sufficient to detect quasi-periodic (QP) photometric varia- each WASP field8 in which L 168-9 was observed. As the photo- tions of L 168-9 due to active regions on the stellar surface metric variations appear to vary nearly periodically, we modeled rotating in and out of view at the stellar rotation period Prot. The the photometry with a Gaussian process (GP) regression model WASP photometry of L 168-9 is shown in Fig.5 along with the and adopted a QP covariance kernel (see Eq. (A.1)) (Angus et al. generalized Lomb-Scargle periodogram (GLSP; Zechmeister & 2018). The covariance function’s periodic timescale was a free Kürster 2009) of the photometry in each of the five WASP observing seasons. It is clear from the GLSPs that a strong peri- 8 At times, L 168-9 appeared within the fields-of-view of multiple odicity exists within the data whose timescale is often 30 days, WASP cameras. ∼ A58, page 7 of 13 A&A 636, A58 (2020) parameter Prot for which the posterior probability density func- In addition to the signal from L 168-9 b, the HARPS RVs tion (PDF) was sampled using a Markov chain Monte Carlo exhibit some power close to Prot and its second harmonic Prot/3. (MCMC) method (see AppendixA). We modeled each binned Although the FAPs of these periodicities over the full frequency WASP light curve with a GP. We adopted a bin size of 1 day to domain are large, they each appear locally as strong periodicities reduce the computation time. In preliminary analyses, we also since most of the power in the HARPS RVs exists at .5 days. tested bin sizes of 0.25, 0.5, and 2 days and found that the recov- The GLSP of the PFS RVs is much more difficult to interpret at ered values of Prot were not very sensitive to this choice, which periodicities in the vicinity of Prot and its first and second har- is probably due to the very large number of points. monics due to strong aliases from the PFS WF. Due to these After sampling the posterior PDFs of the GP hyperparame- effects, it is difficult to discern from the available PFS activity ters, we obtained point estimates for each parameter value based indices (i.e., S-index and Hα) whether or not a coherent activity on the maximum a-posteriori values and 68 percent confidence signal is seen with PFS. Although each of the HARPS and PFS intervals. The resulting mean GP model of the data from each RV time series are significantly affected by sampling aliases, we WASP observing season is depicted in Fig.5 along with the do see evidence for L 168-9 b and rotationally modulated stel- corresponding 1σ confidence interval. Over the five observing lar activity in the RVs we endeavor to mitigate with our adopted seasons, the measured (median) rotation period of L 168-9 was model discussed in Sect. 4.3. P = 29.8 1.3 days. rot ± In principle, the WASP signal could have arisen from any 4.3. Radial velocity and transit model star within the 4800 extraction aperture. However, L 168-9 is the brightest star in the aperture by far. Another concern with Guided by the periodicities in the HARPS and PFS RV time any photometric signal with a period near 30 days is whether series, we proceeded to fit a model to the RVs including the it was affected by moonlight. To check on this possibility, we effects of both stellar activity and the planet. Following numer- searched for modulations in the WASP data of several stars with ous successful applications on both Sun-like (e.g., Haywood a similar brightness within the surrounding 10 arcmin field, but et al. 2014; Grunblatt et al. 2015; Faria et al. 2016; López- we did not find any 30-d signals similar to the one that was Morales et al. 2016; Mortier et al. 2016) and M dwarf stars seen for L 168-9. In any case, the star location is far from (e.g., Astudillo-Defru et al. 2017c; Bonfils et al. 2018; Cloutier the ecliptic, and moonlight contamination is not expected. Fur- et al. 2019a; Ment et al. 2019), we adopted a QP kernel for the thermore, the modulation was sometimes seen at the 15-d first GP as a nonparametric model of the physical processes result- harmonic, which would not be expected for moonlight. We can ing in stellar activity. When used to model RV stellar activity, therefore be confident that the 30-d periodicity in the WASP the QP covariance kernel is often interpreted to model the rota- tional component of stellar activity from active regions on the data arises from L 168-9. The RHK0 from HARPS spectra sup- rotating stellar surface whose lifetimes typically exceed many ports the obtained photometric rotation period, as log(RHK0 ) = 4.562 0.043 (active star) translates into Prot = 22 2 days rotation cycles on M dwarfs (Giles et al. 2017) plus the evo- − ± ± lutionary time scale of the active regions. The corresponding using the RHK0 versus Prot relationship from Astudillo-Defru et al. (2017a). GP hyperparameters are described in detail in AppendixA and include each spectrograph’s covariance amplitudes aHARPS, aPFS, 4.2. Radial velocity periodogram analysis the common exponential timescale λRV, the common coherence parameter ΓRV, and the common periodic timescale PRV equal to A first identification of significant periodicities in the HARPS the stellar rotation period Prot. and PFS RV time series is required in order to develop an accu- The planetary component that is attributed to the transit- rate model of the observed RV variations. In a manner similar to ing planet L 168-9 b was fit to the detrended light curve with our analysis of the WASP photometry, we computed the GLSP a Mandel & Agol(2002) planetary transit model. The detrended of the following HARPS and PFS spectroscopic time series: the TESS light curve was produced by adjusting a QP GP systematic RVs, the window functions (WF), and the S-index, Hα, Hβ, Hγ, model to the photometry alone and with all the transits previ- and the sodium doublet NaD activity indicators. Astudillo-Defru ously removed. The best QP GP model was subtracted to the et al.(2017b) detail how these spectroscopic activity indica- entire TESS data set. The planetary component is modeled by a tors were derived. Each GLSP is shown in Fig.4 along with a Keplerian solution that is parameterized by the planet’s orbital false alarm probability (FAP) that was computed via bootstrap- period Pb, time of midtransit T0, RV semiamplitude K, and the ping with replacement using 104 iterations and normalizing each orbital parameters h = √e cos ω and k = √e sin ω where e and periodogram by its standard deviation. ω are the planet’s orbital eccentricity and argument of peri- Each of the GLSPs of the HARPS and PFS RV time series is astron, respectively. In addition, our RV model contains each dominated by noise and aliases arising from the respective WF. spectrograph’s zero point velocity γHARPS, γPFS and an addi- For example, the HARPS WF contains a forest of peaks with tive scalar jitter sHARPS, sPFS that accounts for any residual jitter comparable FAP from 8 days and extending out toward long that, unlike the stellar activity signal, is not temporally corre- periodicities. Similarly the∼ GLSP of the PFS WF reveals a series lated. The complete RV model therefore contains fourteen model of broad peaks for periodicities &1 day. These features, partic- parameters. ularly those from the PFS WF, have clear manifestations in the To ensure self-consistent planet solutions between the avail- GLSPs of their respective RV and activity indicator time series, able TESS transit data and the RV observations, we simultane- thereby complicating the robust identification of periodocities in ously fit the detrended light curve and the RVs. The common the data. However, strong peaks at the orbital period of L 168-9 b planetary parameters between these two data sets are Pb, T0, ( 1.4 days) are discernible in both RV time series at FAP 0.4% h, and k. The additional model parameters that are required to and∼ 0.3% with HARPS and PFS, respectively. This periodic-∼ model the TESS transit light curve included an additive scalar ∼ ity is not apparent in any of the ancillary activity indicator time jitter sTESS, the baseline flux γTESS, the scaled semimajor axis series, which is expected for a signal originating from an orbiting a/Rs, the planet-star radius ratio rp/Rs, the orbital inclination planet. i, and the nearly-uncorrelated parameters q1 and q2, which are

A58, page 8 of 13 N. Astudillo-Defru et al.: L 168-9 b: a nearby hot terrestrial world

30700

30650 2

30600 0

30550 2 PDCSAP_FLUX [e-/s] 30500 1325 1330 1335 1340 1345 1350

BJD - 2457000 Relative Flux [part-per-thousand]

500

2 0 0

500 2 Relative Flux [ppm] 1000 0.2 0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1.5 1.0 0.5 0.0 0.5 1.0 1.5

Relative Flux [part-per-thousand] Phase Time from mid-transit [hours]

Fig. 6. Time series of TESS data for L 168-9. One fifth of original data are plotted for visualization purposes. Upper panel: detrended TESS light curve. Red vertical bars represent the transits of the planet candidate. Bottom panel: phase folded normalized photometry. Red points correspond to binned data for illustrative purposes. The entire orbital phase is shown to the left, while the right presents the transit phase as well as the best transit model whose parameters come from the Table2.

related to the quadratic limb darkening coefficients u1 and u2 via

2 q1 = (u1 + u2) (2) u1 q2 = (3) 2(u1 + u2) (Kipping 2013). Thus we required eleven model parameters to describe the TESS transit light curve and a total of twenty-one model parameters of the joint RV and light curve data set: Θ = a , a , λ , Γ , P , s , s , s , γ , γ , { HARPS PFS RV RV RV TESS HARPS PFS TESS HARPS γPFS, Pb, T0, K, h, k, a/Rs, rp/Rs, i, q1, q2 . We sampled the posterior PDF of this} 21-dimensional param- eter space using an MCMC sampler. Details on the sampler and the adopted prior distributions on each model parameter are given in AppendixA and Table A.1. Fig. 7. Phase folded radial velocity acquired HARPS (blue points) and PFS (red points) where the best GP model was subtracted. The black 5. Results curve represents the maximum a posteriori model adjusted to the data set. In our analysis of the light curve and radial velocity time series of L 168-9, we evaluated the model presented in Sect. 4.3 on equilibrium temperature between 668 and 965 K assuming a the separate RV data sets obtained with HARPS and PFS as Venus-like and zero bond albedo, respectively. Results for the well as the combined time series. Subtracting the model to entire set of parameters are reported in Table2. Explicitly, we the 46 HARPS radial velocity points reduces the dispersion to report the maximum a posteriori value of each parameter along 1 χ2 = . 3.37 m s− (equivalent to red 3 5), while the dispersion of with its 16th and 84th percentiles, corresponding to a 1σ confi- 1 the 76 PFS residual points gives 4.05 m s− (translating into dence interval. We checked for consistency in our joint analysis 2 χred = 1.5). The dispersion obtained from the 122 HARPS+PFS by performing the analysis for each instrument independently. 1 2 residual points is 3.80 m s− (χred = 2.2). Table2 details the results from this test. We note that HARPS From point estimates of the model parameters Θ from our and PFS results are in agreement within their uncertainties, joint RV plus transit analysis with each of these input data sets, translating into a robust detection of the planetary signal in radial we confirm that L 168-9 b has a radius of 1.39 0.09 R and a velocity data. mass of 4.60 0.56 M , translating into a bulk± mean density⊕ of Figure6 shows the TESS photometry and adjusted transit +2.4 ±3 ⊕ 9.6 1.8 g cm− . The planet is orbiting at 0.02091 0.00024 AU model and Fig.7 illustrates the phase folded radial velocity with from− the parent star; therefore, the hot terrestrial± planet has an the model that best fits the data. A58, page 9 of 13 A&A 636, A58 (2020)

Table 3. Prospects of atmospheric characterization of confirmed terrestrial planets including L 168-9 b.

Star ID Rp Mp P a Teff Teq Tday JKs R? M? TSM ESM Ref. units [R ][M ] [days] [AU] [K] [K] [K] [mag] [mag] [R ][M ] ⊕ ⊕ L 168-9 b 1.39 4.39 1.401 0.021 3743 963.01 1059.31 7.941 7.0819 0.60 0.62 8.025 9.692 LHS 3844 b 1.30 – 0.463 0.006 3036 805.90 886.49 10.046 9.145 0.19 0.15 – 29.004 Kr2019 GJ 1132 b 1.13 1.66 1.629 0.015 3270 590.61 649.67 9.245 8.322 0.21 0.18 31.166 9.872 Bo2018 L 98-59 c 1.35 2.17 3.690 0.032 3412 515.31 566.84 7.933 7.101 0.31 0.31 29.168 6.696 Cl2019b LTT 1445A b 1.38 2.20 5.359 0.038 3335 433.34 476.68 7.29 6.5 0.28 0.26 44.976 6.382 Wi2019 TRAPPIST-1 b 1.09 1.02 1.511 0.011 2559 402.38 442.62 11.4 10.3 0.12 0.08 36.914 4.007 Gi2017 LHS 1140 c 1.28 1.81 3.778 0.027 3216 436.43 480.07 9.612 8.821 0.21 0.18 25.225 3.401 Me2019

Notes. TSM and ESM correspond to transmission and emission spectroscopy metrics, respectively. References. Kr2019 – Kreidberg et al.(2019); Bo2018 – Bonfils et al.(2018); Cl2019b – Cloutier et al.(2019b); Wi2019 – Winters et al.(2019b); Gi2017 – Gillon et al.(2017); Me2019 – Ment et al.(2019).

6. Discussion and conclusions 4.0 100% Fe

L 168-9 b adds to the family of small (<2 R ) transiting plan- 50% Fe - 50% MgSiO3 ⊕ 3.5 ets around bright (J < 8 mag) stars with mass measurements 100% MgSiO3 and contributes to the completion of the TESS level one sci- 3.0 50% H2O - 50% MgSiO3 ence requirement to detect and measure the masses of 50 small 100% H2O planets. In particular, L 168-9 b is one of fourteen9 likely rocky planets without primordial hydrogen-helium envelopes – that 2.5 R is, it has a radius <1.8 R – for which the mass has been / measured with an uncertainty⊕ that is smaller than 33%. Thus, R 2.0 our result represents progress toward the understanding of the transition between super-Earths and mini-Neptunes as previ- 1.5 ously reported in the radii of planets (e.g., Fulton et al. 2017; Cloutier & Menou 2019) but, here, it includes information about 1.0 mass. Figure8 shows the mass-radius diagram that is centered in 0.5 the sub-Earth to mini- regime. With about twice the 1 10 Earth average density, L 168-9 b bulk density is compatible with M/M a with an iron core (50%) surrounded by a mantle of silicates (50%). In this diagram the detected planet Fig. 8. Mass-radius diagram showing L 168-9 b (red circle) in the con- is located in an interesting place: for masses lower than that text of known . The transparency of each point is proportional of L 168-9 b, the great majority of planets are consistent with to its associated mass uncertainty. Error bars correspond to 1σ uncer- tainties. Different models for the bulk composition are plotted where a 50% Fe–50% MgSiO3 or 100% MgSiO3 bulk composition, the legend details the fraction of iron, silicates, and/or water for each while there is a great diversity in density for higher planetary color-coded curve. masses. Being one of the densest planets for masses greater than 4 M , L 168-9 b can help to define the mass limits of the rocky ⊕ planets population. shorter than 10 days from Dressing & Charbonneau(2015) and Good targets for atmospheric characterization with trans- combining this with the transit probability of such planets (about mission and emission spectroscopy are those transiting nearby, a couple % to 20%), there would be a few and up to about < bright stars (J 8 mag). There are currently 11 small planets twenty such planets.∼ detected transiting a bright star, according to the NASA Exo- 10 The measured properties of L 168-9 b and its host star planet Archive , two of them orbit M dwarfs. Overall, there make it a promising target for the atmospheric characteriza- would not be a large number of small, transiting planets around tion of a terrestrial planet via either transmission or emission nearby, bright M dwarfs. There are roughly 200 M dwarfs within spectroscopy measurements with JWST (Morley et al. 2017; the 25 pc solar neighborhood with J < 8 mag, about 2/3 of which Kempton et al. 2018) and/or thermal analysis to are single stars (Winters et al. 2015, 2019a). Considering the infer the absence of a thick atmosphere (e.g., Seager & Deming occurrence rate of small planets with an orbital period that is 2009). Its transmission and emission spectroscopy metrics from Kempton et al.(2018) are reported in Table3 and they are 9 L 98-59 bc (Cloutier et al. 2019b), GJ 357 b (Luque et al. 2019), compared to other confirmed transiting terrestrial planets with HD 15337 b (Dumusque et al. 2019), HD 213885 b (Espinoza et al. known masses that are of interest for atmospheric character- 2020), GJ 9827 b (Rice et al. 2019), K2-265 b (Lam et al. 2018), K2-141 b (Barragán et al. 2018), K2-229 b (Santerne et al. 2018), ization. Based on this assessment, L 168-9 b is an excellent HD 3167 b (Gandolfi et al. 2017), K2-106 b (Guenther, E. W. et al. 2017), candidate for emission spectroscopy or for detecting the plan- TRAPPIST-1 fh (Wang et al. 2017), HD 219134 b (Motalebi et al. 2015). etary day-side phase curve as was recently done for the similar 10 https://exoplanetarchive.ipac.caltech.edu/ planet, LHS 3844 b (Kreidberg et al. 2019).

A58, page 10 of 13 N. Astudillo-Defru et al.: L 168-9 b: a nearby hot terrestrial world

Acknowledgements. N.A.-D. acknowledges the support of FONDECYT project Cloutier, R., Astudillo-Defru, N., Bonfils, X., et al. 2019b, A&A, 629, A111 3180063. J.K.T. acknowledges that support for this work was provided by NASA Collins, K. A., Kielkopf, J. F., Stassun, K. G., & Hessman, F. V. 2017, AJ, 153, through Hubble Fellowship grant HST-HF2-51399.001 awarded by the Space 77 Telescope Science Institute, which is operated by the Association of Universi- Crane, J. D., Shectman, S. A., & Butler, R. P. 2006, Proc. SPIE, 6269, 626931 ties for Research in Astronomy, Inc., for NASA, under contract NAS5-26555. Crane, J. D., Shectman, S. A., Butler, R. P., Thompson, I. B., & Burley, G. S. R.B. acknowledges support from FONDECYT Post-doctoral Fellowship Project 2008, Proc. SPIE, 7014, 701479 3180246, and from the Millennium Institute of Astrophysics (MAS). X.B. and Crane, J. D., Shectman, S. A., Butler, R. P., et al. 2010, Proc. SPIE, 7735, 773553 J.M-A. acknowledge funding from the European Research Council under the Cutri, R. M., Skrutskie, M. F., van Dyk, S., et al. 2003, VizieR Online Data ERC Grant Agreement n. 337591-ExTrA. X.D., X.B., T.F., and L.M. acknowl- Catalog: II/246 edge the support by the French National Research Agency in the framework of Cutri, R. M., Wright, E. L., Conrow, T., et al. 2013, VizieR Online Data Catalog: the Investissements d’Avenir program (ANR-15-IDEX-02), through the funding II/328 of the “Origin of Life” project of the Univ. Grenoble-Alpes. L.M. acknowedges de Zeeuw, T., Tamai, R., & Liske, J. 2014, The Messenger, 158, 3 the support of the Labex OSUG@2020 (Investissements d’avenir – ANR10 Dressing, C. D., & Charbonneau, D. 2015, ApJ, 807, 45 LABX56). J.R.M. acknowledges CAPES, CNPq and FAPERN brazilian agen- Dumusque, X., Turner, O., Dorn, C., et al. 2019, A&A, 627, A43 cies. This work was supported by FCT/MCTES through national funds and Espinoza, N., Brahm, R., Henning, T., et al. 2020, MNRAS, 491, 2982 by FEDER – Fundo Europeu de Desenvolvimento Regional through COM- Faria, J. P., Haywood, R. D., Brewer, B. J., et al. 2016, A&A, 588, A31 PETE2020 – Programa Operacional Competitividade e Internacionalização by Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. 2013, PASP, 125, these grants: UID/FIS/04434/2019; PTDC/FIS-AST/32113/2017 & POCI-01- 306 0145-FEDER-032113; PTDC/FIS-AST/28953/2017 & POCI-01-0145-FEDER- Fulton, B. J., Petigura, E. A., Howard, A. W., et al. 2017, AJ, 154, 109 028953. T.H. acknowledges support from the European Research Council under Gaia Collaboration 2018, VizieR Online Data Catalog: I/345 the Horizon 2020 Framework Program via the ERC Advanced Grant Origins Gaidos, E., Mann, A. W., Lépine, S., et al. 2014, MNRAS, 443, 2561 83 24 28. R.E.M. acknowledges support by the BASAL Centro de Astrofísica y Gandolfi, D., Barragán, O., Hatzes, A. P., et al. 2017, AJ, 154, 123 Tecnologías Afines (CATA) and FONDECYT 1190621. A.J. acknowledges sup- Gardner, J. P., Mather, J. C., Clampin, M., et al. 2006, Space Sci. Rev., 123, 485 port from FONDECYT project 1171208 and by the Ministry for the Economy, Giles, H. A. C., Collier Cameron, A., & Haywood, R. D. 2017, MNRAS, 472, Development, and Tourism’s Programa Iniciativa Científica Milenio through 1618 grant IC 120009, awarded to the Millennium Institute of Astrophysics (MAS). Gillon, M., Triaud, A. H. M. J., Demory, B.-O., et al. 2017, Nature, 542, 456 J.G.W. is supported by a grant from the John Templeton Foundation. The opin- Goodman, J., & Weare, J. 2010, Commun. Appl. Math. Comput. Sci., 5, 65 ions expressed here are those of the authors and do not necessarily reflect the Grunblatt, S. K., Howard, A. W., & Haywood, R. D. 2015, ApJ, 808, 127 views of the John Templeton Foundation. Work of J.N.W. was partly funded by Guenther, E. W., Barragán, O., Dai, F., et al. 2017, A&A, 608, A93 the Heising-Simons Foundation. The authors would like to acknowledge Zachary Hauschildt, P. H., Allard, F., & Baron, E. 1999, ApJ, 512, 377 Hartman for his help conducting the Gemini-South/DSSI observations. Some Haywood, R. D., Collier Cameron, A., Queloz, D., et al. 2014, MNRAS, 443, of the work here is based on observations obtained at the Gemini Observatory, 2517 which is operated by the Association of Universities for Research in Astron- Henden, A. A., Templeton, M., Terrell, D., et al. 2016, VizieR Online Data omy, Inc., under a cooperative agreement with the NSF on behalf of the Gemini Catalog: II/336 partnership: the National Science Foundation (United States), National Research Henry, G. W., Marcy, G. W., Butler, R. P., & Vogt, S. S. 2000, ApJ, 529, L41 Council (Canada), CONICYT (Chile), Ministerio de Ciencia, Tecnología e Inno- Høg, E., Fabricius, C., Makarov, V. V., et al. 2000, A&A, 355, L27 vación Productiva (Argentina), Ministério da Ciência, Tecnologia e Inovação Horch, E. P., Veillette, D. R., Baena Gallé, R., et al. 2009, AJ, 137, 5057 (Brazil), and Korea Astronomy and Space Science Institute (Republic of Korea). Horch, E. P., Gomez, S. C., Sherry, W. H., et al. 2011, AJ, 141, 45 This work makes use of observations from the LCOGT network. Funding for the Horch, E. P., Howell, S. B., Everett, M. E., & Ciardi, D. R. 2012, AJ, 144, 165 TESS mission is provided by NASA’s Science Mission directorate. We acknowl- Howell, S. B., Everett, M. E., Sherry, W., Horch, E., & Ciardi, D. R. 2011, AJ, edge the use of public TESS Alert data from pipelines at the TESS Science Office 142, 19 and at the TESS Science Processing Operations Center. This research has made Jenkins, J. M., Twicken, J. D., McCauliff, S., et al. 2016, SPIE Conf. Ser., 9913, use of the Exoplanet Follow-up Observation Program website, which is operated 99133E by the California Institute of Technology, under contract with the National Aero- Jensen, E. 2013, Astrophysics Source Code Library [ascl:1306.007] nautics and Space Administration under the Exoplanet Exploration Program. Kempton, E. M.-R., Bean, J. L., Louie, D. R., et al. 2018, PASP, 130, 114401 Resources supporting this work were provided by the NASA High-End Com- Kipping, D. M. 2013, MNRAS, 435, 2152 puting (HEC) Program through the NASA Advanced Supercomputing (NAS) Kreidberg, L., Koll, D. D. B., Morley, C., et al. 2019, Nature, 573, 87 Division at Ames Research Center for the production of the SPOC data products. Lam, K. W. F., Santerne, A., Sousa, S. G., et al. 2018, A&A, 620, A77 This paper includes data collected by the TESS mission, which are publicly avail- Lo Curto, G., Pepe, F., Avila, G., et al. 2015, The Messenger, 162, 9 able from the Mikulski Archive for Space Telescopes (MAST). This research has López-Morales, M., Haywood, R. D., Coughlin, J. L., et al. 2016, AJ, 152, 204 made use of the NASA Exoplanet Archive, which is operated by the California Lovis, C., & Pepe, F. 2007, A&A, 468, 1115 Institute of Technology, under contract with the National Aeronautics and Space Luque, R., Pallé, E., Kossakowski, D., et al. 2019, A&A, 628, A39 Administration under the Exoplanet Exploration Program. Mandel, K., & Agol, E. 2002, ApJ, 580, L171 Mann, A. W., Dupuy, T., Kraus, A. L., et al. 2019, ApJ, 871, 63 Mayor, M., Pepe, F., Queloz, D., et al. 2003, The Messenger, 114, 20 References Mazeh, T., Naef, D., Torres, G., et al. 2000, ApJ, 532, L55 Ment, K., Dittmann, J. A., Astudillo-Defru, N., et al. 2019, AJ, 157, 32 Allard, F. 2014, IAU Symp., 299, 271 Moór, A., Ábrahám, P., Derekas, A., et al. 2006, ApJ, 644, 525 Anglada-Escudé, G., & Butler, R. P. 2012, ApJS, 200, 15 Morley, C. V., Kreidberg, L., Rustamkulov, Z., Robinson, T., & Fortney, J. J. Angus, R., Morton, T., Aigrain, S., Foreman-Mackey, D., & Rajpaul, V. 2018, 2017, ApJ, 850, 121 MNRAS, 474, 2094 Mortier, A., Faria, J. P., Santos, N. C., et al. 2016, A&A, 585, A135 Astudillo-Defru, N., Delfosse, X., Bonfils, X., et al. 2017a, A&A, 600, A13 Motalebi, F., Udry, S., Gillon, M., et al. 2015, A&A, 584, A72 Astudillo-Defru, N., Forveille, T., Bonfils, X., et al. 2017b, A&A, 602, A88 Muirhead, P. S., Dressing, C. D., Mann, A. W., et al. 2018, AJ, 155, 180 Astudillo-Defru, N., Díaz, R. F., Bonfils, X., et al. 2017c, A&A, 605, L11 Neves, V., Bonfils, X., Santos, N. C., et al. 2014, A&A, 568, A121 Baranne, A., Queloz, D., Mayor, M., et al. 1996, A&AS, 119, 373 Pollacco, D. L., Skillen, I., Collier Cameron, A., et al. 2006, PASP, 118, 1407 Barragán, O., Gandolfi, D., Dai, F., et al. 2018, A&A, 612, A95 Queloz, D., Henry, G. W., Sivan, J. P., et al. 2001, A&A, 379, 279 Batalha, N. E., Lewis, T., Fortney, J. J., et al. 2019, ApJ, 885, L25 Rice, K., Malavolta, L., Mayo, A., et al. 2019, MNRAS, 484, 3731 Bayo, A., Rodrigo, C., Barrado Y Navascués, D., et al. 2008, A&A, 492, 277 Ricker, G. R., Winn, J. N., Vanderspek, R., et al. 2015, J. Astron. Telesc. Instrum. Beichman, C. A., Bryden, G., Stapelfeldt, K. R., et al. 2006, ApJ, 652, 1674 Syst., 1, 014003 Bonfils, X., Almenara, J.-M., Cloutier, R., et al. 2018, A&A, 618, A142 Santerne, A., Brugger, B., Armstrong, D. J., et al. 2018, Nat. Astron., 2, 393 Brown, T. M., Baliber, N., Bianco, F. B., et al. 2013, PASP, 125, 1031 Seager, S., & Deming, D. 2009, ApJ, 703, 1884 Butler, R. P., Marcy, G. W., Williams, E., et al. 1996, PASP, 108, 500 Smith, J. C., Stumpe, M. C., Van Cleve, J. E., et al. 2012, PASP, 124, 1000 Charbonneau, D., Brown, T. M., Latham, D. W., & Mayor, M. 2000, ApJ, 529, Stassun, K. G., & Torres, G. 2016, AJ, 152, 180 L45 Stassun, K. G., & Torres, G. 2018, ApJ, 862, 61 Ciardi, D. R., Beichman, C. A., Horch, E. P., & Howell, S. B. 2015, ApJ, 805, 16 Stassun, K. G., Collins, K. A., & Gaudi, B. S. 2017, AJ, 153, 136 Cloutier, R., & Menou, K. 2019, AAS J., submitted [arXiv:1912.02170] Stassun, K. G., Corsaro, E., Pepper, J. A., & Gaudi, B. S. 2018a, AJ, 155, 22 Cloutier, R., Astudillo-Defru, N., Doyon, R., et al. 2017, A&A, 608, A35 Stassun, K. G., Oelkers, R. J., Pepper, J., et al. 2018b, AJ, 156, 102 Cloutier, R., Astudillo-Defru, N., Doyon, R., et al. 2019a, A&A, 621, A49 Stumpe, M. C., Smith, J. C., Catanzarite, J. H., et al. 2014, PASP, 126, 100

A58, page 11 of 13 A&A 636, A58 (2020)

Sullivan, P. W., Winn, J. N., Berta-Thompson, Z. K., et al. 2015, ApJ, 809, 77 19 Cerro Tololo Inter-American Observatory, Casilla 603, La Serena, Teske, J. K., Shectman, S. A., Vogt, S. S., et al. 2016, AJ, 152, 167 Chile Teske, J. K., Ciardi, D. R., Howell, S. B., Hirsch, L. A., & Johnson, R. A. 2018, 20 Department of Terrestrial Magnetism, Carnegie Institution for AJ, 156, 292 Science, 5241 Broad Branch Road NW, Washington DC 20015, USA Tokovinin, A. 2018, PASP, 130, 035002 21 American Association of Observers, 49 Bay State Twicken, J. D., Catanzarite, J. H., Clarke, B. D., et al. 2018, PASP, 130, 064502 Road, Cambridge, MA 02138, USA Wang, S., Wu, D.-H., Barclay, T., & Laughlin, G. P. 2017, ApJ, submitted 22 [arXiv:1704.04290] Department of Physics, and Kavli Institute for Astrophysics and Winters, J. G., Henry, T. J., Lurie, J. C., et al. 2015, AJ, 149, 5 Space Research, Massachusetts Institute of Technology, Cambridge, Winters, J. G., Henry, T. J., Jao, W.-C., et al. 2019a, AJ, 157, 216 MA, USA Winters, J. G., Medina, A. A., Irwin, J. M., et al. 2019b, AJ, 158, 152 23 Physics & Astronomy Department, University of Kansas, Lawrence, Zechmeister, M., & Kürster, M. 2009, A&A, 496, 577 KS 66044, USA Ziegler, C., Law, N. M., Baranec, C., et al. 2018, AJ, 155, 161 24 Universidad de Buenos Aires, Facultad de Ciencias Exactas y Ziegler, C., Tokovinin, A., Briceno, C., et al. 2020, AJ, 159, 19 Naturales. Buenos Aires, Argentina 25 CONICET – Universidad de Buenos Aires, Instituto de Astronomía y Física del Espacio (IAFE), Buenos Aires, Argentina 26 Department of Physics and Astronomy, University of New Mexico, 1 Departamento de Matemática y Física Aplicadas, Universidad Albuquerque, NM, USA Católica de la Santísima Concepción, Alonso de Rivera 2850, 27 Science Institute, 3700 San Martin Drive, Concepción, Chile Baltimore, MD 21218, USA e-mail: [email protected] 28 European Southern Observatory, Alonso de Córdova 3107, Vitacura, 2 Center for Astrophysics | Harvard & Smithsonian, 60 Garden Street, Región Metropolitana, Chile Cambridge, MA 02138, USA 29 Universidad de Concepción, Departamento de Astronomía, Casilla 3 Department of Astronomy & Astrophysics, University of Toronto, 160-C, Concepción, Chile 50 St. George Street, Toronto, ON, M5S 3H4, Canada 30 Physics Department and Tsinghua Centre for Astrophysics, Tsinghua 4 Observatories of the Carnegie Institution for Science, 813 Santa University, Beijing 100084, PR China Barbara Street, Pasadena, CA 91101, USA 31 Centre for Astrophysics, University of Southern Queensland, 5 Center of Astro-Engineering UC, Pontificia Universidad Católica de Toowoomba, QLD 4350, Australia Chile, Av. Vicuña Mackenna 4860, 7820436 Macul, Santiago, Chile 32 Max-Planck-Institut fur Astronomie, Künigstuhl 17, 69117 6 Instituto de Astrofísica, Facultad de Física, Pontificia Universidad Heidelberg, Germany Católica de Chile, Av. Vicuña Mackenna 4860, 7820436 Macul, 33 Department of Physics, Southern Connecticut State University, 501 Santiago, Chile Crescent Street, New Haven, CT 06515, USA 7 Millennium Institute of Astrophysics, Santiago, Chile 34 Campo Catino Astronomical Observatory, Regione Lazio, Guarcino 8 Astrophysics Group, Keele University, Staffordshire, ST5 5BG, UK (FR) 03010, Italy 9 Kavli Institute for Astrophysics and Space Research, Massachusetts 35 Departamento de Astronomía, Universidad de Chile, Camino El Institute of Technology, Cambridge, MA 02139, USA Observatorio 1515, Las Condes, Santiago, Chile 10 Department of Earth, Atmospheric, and Planetary Sciences, 36 Facultad de Ingeniería y Ciencias, Universidad Adolfo Ibáñez, Av. Massachusetts Institute of Technology, Cambridge, MA 02139, Diagonal las Torres 2640, Peñalolén, Santiago, Chile USA 37 Department of Physics and Astronomy, University of Louisville, 11 Department of Astrophysical Sciences, Princeton University, Louisville, KY 40292, USA Princeton, NJ 08544, USA 38 Department of Physics and Astronomy, The University of North 12 NASA Ames Research Center, Moffett Field, CA 94035, USA Carolina at Chapel Hill, Chapel Hill, NC 27599-3255, USA 13 Department of Physics and Astronomy, Vanderbilt University, 39 Departamento de Física, Universidade Federal do Rio Grande do Nashville, TN 37235, USA 14 Norte, 59078-970 Natal, RN, Brazil Dunlap Institute for Astronomy and Astrophysics, University of 40 Georgia State University, Department of Physics & Astronomy, 25 Toronto, Ontario M5S 3H4, Canada 15 Park Place NE #605, USA Univ. Grenoble Alpes, CNRS, IPAG, 38000 Grenoble, France 41 Instituto de Astrofísica e Ciências do Espaço, Universidade do Porto, 16 Department of Physics, University of Warwick, Gibbet Hill Road, CAUP, Rua das Estrelas, PT4150-762 Porto, Portugal Coventry CV4 7AL, UK 42 Departamento de Física e Astronomia, Faculdade de Ciências, 17 Institut de Recherche sur les Exoplanètes, Département de Physique, Universidade do Porto, Porto, Portugal Université de Montréal, Montréal QC, H3C 3J7, Canada 43 SETI Institute, Moffett Field, CA 94035, USA 18 Observatoire de Genève, Université de Genève, 51 ch. des Maillettes, 44 Center for Space Research, MIT, 37-414, Cambridge, MA 02139, 1290 Sauverny, Switzerland USA

A58, page 12 of 13 N. Astudillo-Defru et al.: L 168-9 b: a nearby hot terrestrial world

Appendix A: Gaussian process model Table A.1. L 168-9 model parameter priors. Gaussian process (GP) regression is widely used in the exo- planet community as a nonparametric Bayesian approach to Parameter Prior model temporally correlated stellar activity signals in RV data. Photometry model These rotationally-modulated activity signals prohibit the accu- Baseline flux, γ ( 0.1, 0.1) rate measurement of planetary parameters and often produce in TESS U − Limb darkening coefficient, u (0.1, 0.4) planetary false positives. 1 U Limb darkening coefficient, u2 (0.25, 0.55) Here, we modeled the RV activity signals of L 168-9 as a U 2 TESS additive jitter, s (10− , 1) stochastic process whose temporal evolution is well-described by TESS J a quasi-periodic covariance kernel. A GP with a quasi-periodic RV model 1 covariance kernel k(ti, t j) is included in our joint model describ- HARPS zero point velocity, γ0,HARPS [m s− ] ( 1, 1) 1 U − ing the data RV and transit data and takes the following form: PFS zero point velocity, γ0,PFS [m s− ] ( 1, 1) 1 U − " 2 !# ln HARPS covariance amplitude, ln (aHARPS/m s− ) ( 5, 10) 2 (ti t j) 2 2 π ti t j 1 U − , = − Γ | − | ln PFS covariance amplitude, ln (aPFS/m s− ) ( 5, 10) k(ti t j) a exp 2 sin (A.1) U − − 2λ − PGP ln RV exponential time scale, ln (λRV/day) From Hα training ln RV coherence, ln (Γ ) From Hα training and is described by the covariance amplitude a, the exponen- RV ln RV periodic timescale, ln (PRV/day) From Hα training tial evolutionary timescale λ, the coherence Γ, and the periodic 1 2 HARPS additive jitter, sHARPS [m s− ] (10− , 10) timescale PGP. 1 J 2 PFS additive jitter, s [m s− ] (10− , 10) As usual in the Bayesian context, prior probability density PFS J functions (PDF) for the hyperparameters a, λ, Γ, P are listed L 168-9 b (TOI-134.01) { GP} in Table A.1 for the multiple GPs that were applied in our data Orbital period, Pb [days] (1.370, 1.405) U analysis. The posterior PDF were sampled through a MCMC Time of midtransit, T0,b [BJD 2 457 000] (338, 1342) − U (Goodman & Weare 2010); in particular, we used the emcee Scaled semimajor axis, a/Rs (7.542, 0.27) G ensemble sampler (Foreman-Mackey et al. 2013). The sampling Planet-star radius ratio, rp/Rs (0, 0.1) U of the joint posterior was made with the Gaussian ln likelihood Orbital inclination, i [deg] (75, 105) 1 U 2 function given by Semiamplitude, Kb [m s− ] (10− , 10) J h = √e cos ω ( 1, 1) 1   b b b U − ln = yT K y + ln detK + N ln 2π (A.2) k = √e sin ω ( 1, 1) L −2 · · b b b U − where y is the vector of N measurements taken at times t = t1, t2,..., tN and the N N covariance matrix K is given by { } × with δi j being the Kronecker delta that adds the measurement 2 2 Ki j = k(ti, t j) + δi j[σ(ti) + s ] (A.3) uncertainties σ to the diagonal elements of K and includes an additive jitter factor s.

Appendix B: Spectroscopic data Table B.1. HARPS radial velocity time series and spectroscopic activity indicators for L 168-9.

BJD RV σRV Hα σHα Hβ σHβ Hγ σHγ NaD σNaD S σS 1 1 [-2 450 000] [ms− ] [ms− ] 4664.954608 29 781.65 3.48 0.06224 0.00025 0.04679 0.00052 0.10087 0.00131 0.01232 0.00020 1.616 0.068 4991.927461 29 786.04 2.95 0.06074 0.00021 0.04632 0.00044 0.10385 0.00114 0.01172 0.00016 2.167 0.073 8367.523380 29 769.97 1.37 0.05962 0.00011 0.04606 0.00019 0.10582 0.00052 0.01166 0.00007 1.892 0.018 8367.551447 29 773.09 1.19 0.05928 0.00010 0.04534 0.00016 0.10573 0.00045 0.01158 0.00006 1.826 0.014 8367.623921 29 772.57 1.20 0.05944 0.00010 0.04562 0.00017 0.10639 0.00046 0.01164 0.00006 1.868 0.014

Notes. The full version is available at the CDS.

Table B.2. PFS radial velocity time series and spectroscopic activity indicators.

BJD RV σRV Hα S [ 2 450 000] [ms 1] [ms 1] − − − 8409.51799 8.80 1.74 1.293 0.05807 8409.53206 4.10 1.74 1.334 0.05813 8409.60491 3.11 1.71 1.285 0.05879 8409.61960− 4.80 1.75 1.340 0.05901 8409.67495 8.23 2.09 1.311 0.05976

Notes. The full version is available at the CDS.

A58, page 13 of 13